You've persuaded me to get the book, I haven't read it yet but I listened to the podcast. I have a few errant thoughts, not exhaustive, it's what I remember.

You've persuaded me to get the book, I haven't read it yet but I listened to the podcast. I have a few errant thoughts, not exhaustive, it's what I remember. 1. She described the first diagram of a home-automation device with 2 nodes, the device and the device owner. She made the case that in that schematic, the owner's privacy, autonomy, agency is respected. She made the contrast that Nest requires thousands of agreements w/r/t data collected. I understood this to mean she prefers something like this model, no 3rd party interaction. This is impossible if the vast majority of Nest users (as an example) cannot program Nest to deliver temperature, on/off switches to their phone. You can see the problem, the world is increasingly software driven and despite the "everyone should learn to code" homily, this is not so and doesn't look like it's going to be so in our future. As a corollary to (1), I wonder if this is an inherent attribute of the software which we write to day. If we are to write software for a device which we will never interact with again, or have very few opportunity to interact with it for repair purposes (eg, software sent to the moon or Mars), it would be a different kind of software written with a different maintenance model. We build remote capabilities into devices (eg Nest) because we ship an MVP, first to market, and we build the reach-back into it to fix it afterward. Its first purpose is not to collect data, I don't think. The data collection mechanisms and purposes accreted over time. 2. She segued into the creation of the mobile phone as the next step in creating the predictive product, after the click-through ads. "Let us give you this device, maybe a smartphone ..." Who is this "us"? The smartphone was an iterative product with an inflection point. Personal organizers, navigation systems, existed before the smartphone and did not find market viability (Newton, Palm, Garmin), not the way the smartphone did. The smartphone is what it is because it fulfilled consumers' needs and customers demanded it. This line of deduction seems backward. 3. She said "predictive products" (I forget the exact words she used, things that predicts human actions) should not exist. Even the interviewer said that's a radical prescription! Humans have been doing this, wanted to do this forever. We want to use what we know to "nudge" behaviors in a certain direction, there's actually a thing in behavioral economic called "nudging", nudge people to invest in 401K, eat their veg, not fall into addiction and crime. I'm not sure how she proposes to do away with this. She will say I'm committing a category error, by conflating the predictive nudge with capitalism. I'm not sure how that can be decoupled, because as a society, capitalism is how we allocate values, and money is a proxy for value. I'm not saying the allocation is always correctly coupled or aligned but if a behavior is valuable, then the nudge toward that behavior is valuable and the information that enables the delivery of the nudge is valuable. That's my immediate reaction as I listened, the book will have a more nuanced discussion I'm sure. I am willing to be persuaded but I'm forever skeptical of social scientists' conclusions about the workings, intents, and outcomes of tech decisions.